AWS Public Sector Blog

Category: Amazon Bedrock

Generative AI ROI and governance framework for space using Amazon Bedrock

Mission AI ROI strategy framework for space using Amazon Bedrock

This post introduces the Governed Intelligence Framework (GI Framework), giving you a defensible path from AI spending to quantifiable value without compromising governance rigor. Built on Amazon Bedrock by Amazon Web Services (AWS) as the implementation solution, the framework rests on a single argument: ROI, governance, and token optimization form a reinforcing system. None of them deliver results in isolation.

How Muzzy Lane and CSU Global built a crisis intervention roleplay assessment with Amazon Bedrock

Colorado State University Global (CSU Global), a 100% online, accredited nonprofit university, set out to take on this opportunity by collaborating with Muzzy Lane, an educational technology company that supports instructional designers to design and deliver roleplay assessments directly in their learning management system (LMS). Muzzy Lane’s Author platform is powered by Amazon Bedrock, which supports its AI capabilities and protects customer data. By using Author, CSU Global designed its high-stakes roleplay assessments for its Bachelor of Science (BS) in Human Services program.

Building an environmental intelligence platform on AWS with open data

In this blog, learn how we built a working prototype of a serverless environmental intelligence platform, where nontechnical users ask questions in plain language and receive answers drawn from open data, published models, and satellite observations. The platform is built on four pillars, as outlined in the following table.

$Fewer than 5 percent of radiologists review their own billing codes, compared to roughly 90 percent of primary care physicians. That gap has consequences: Over the past decade, it has contributed to nearly 50 percent reimbursement losses in radiology. In radiology, radiologists' adjacent personnel must translate every image read into standardized International Classification of Diseases, 10th Revision (ICD-10) codes that drive billing, follow-ups, and quality reporting. When those codes are unverified or poorly documented, the results are billing delays, claim denials, and delayed patient care. The University of Miami Health System (UHealth) partnered with AWS and AWS Partner Quantiphi to build a generative AI coding solution on Amazon Bedrock called Hurricode. The solution puts radiologists back in the loop, achieving approximately 92 percent coding accuracy, projecting a 34 percent revenue increase, and reducing manual effort by roughly 40 percent. The University of Miami Leonard M. Miller School of Medicine is Florida's first medical school and home to the Sylvester Comprehensive Cancer Center, the highest-ranked cancer center in Florida for cancer care in 2026. The Department of Radiology performs over 1 million procedures annually across eight specialized divisions. Quantiphi is an AWS Premier Tier Services Partner and AWS 2025 Public Sector Global Generative AI Consulting Partner of the Year. When manual coding can't keep up The UHealth radiology team faced a set of interconnected problems. The sheer volume of manual reporting fueled clinician fatigue. The department's legacy natural language processing (NLP) coding tool reached only 58–77 percent accuracy. It lacked the precision, scalability, and speed that modern diagnostic workflows demand, producing inconsistent ICD-10 code mapping and driving medical necessity denials. A lack of transparency compounded the inconsistency. The team needed defensible code identification in the form of clear, understandable rationales behind every ICD-10 prediction. Without it, clinicians couldn't trust or validate the output, and the department couldn't meet its compliance and audit requirements. UHealth serves a complex oncologic and tertiary care population where actionable incidental findings (AIFs) such as lung nodules, pulmonary emboli, and fractures demand timely follow-up. This makes accurate, well-documented coding a patient safety issue as much as a financial one. Building a radiologist-in-the-loop workflow on AWS To solve these challenges, UHealth and Quantiphi designed a solution with radiologists, studying their exact workflow to understand where AI could reduce friction without disrupting clinical judgment. The result is Hurricode, a generative AI–powered coding assistant built on Amazon Bedrock. Hurricode uses Amazon Bedrock to generate ICD-10 code suggestions with associated reasoning from transcribed radiology reports. Radiologists review and confirm the codes through a custom interface, a self-attestation process that takes less than 30 seconds per study. This radiologist-in-the-loop approach maintains clinical-grade accuracy while reducing manual effort by approximately 40 percent. The solution also accelerates upstream insurer verification (pre-authorization) for recommended further imaging, which occurs in 11–27 percent of advanced imaging studies. By surfacing accurate, well-documented codes earlier in the workflow, Hurricode helps streamline the pre-authorization process. The project began with a strategic assessment, roadmap, and proof of concept (PoC). The pipeline ingests, pre-processes, and passes radiology reports and the ICD-10 code directory to a fine-tuned model. Hurricode is the first solution to flag pertinent negative findings (PNFs)—conditions that have been ruled out, such as bleeds or fractures—improving documentation quality and supporting more complete clinical records. Future plans include surfacing AIFs for tracking, scheduling, and pre-authorization through the University of Miami No Findings Left Behind™ provenance network. Figure 1: ICD-10 Coding Automation Workflow for Radiology Reports How the solution comes together on AWS The following AWS services power the Hurricode solution: Amazon Bedrock — Foundation models including Amazon Titan Text Embeddings and Anthropic Claude for phrase extraction, code generation, and AIF identification Amazon SageMaker — Development and fine-tuning of the embedding model Amazon OpenSearch Service — Stores vectors for semantic search AWS Lambda — Orchestrates data ingestion and RAG (Retrieval Augmented Generation) pipeline Amazon EC2 — Additional compute Amazon S3 — Stores raw inputs, processed text, and web assets Amazon DynamoDB and Amazon RDS — Manage application and structured metadata Amazon CloudFront and Amazon Cognito — Deliver web application more securely, integrated with University of Miami single sign-on Figure 2: AWS Cloud Architecture for AI-Powered Medical Coding System Clinical-grade results at scale Hurricode has delivered measurable improvements across coding accuracy, operational efficiency, and clinical documentation: Approximately 92 percent coding accuracy, up from 58–77 percent Approximately 40 percent reduction in manual effort Radiologists spend less than 30 seconds per advanced imaging study Projected approximately 34 percent revenue increase through optimized revenue cycle management and advanced authorization Additional upside from CMS Quality Payment Program quality metrics Physician attestation improves documentation of PNFs in more than 50 percent of patients Projected 10–25 percent uplift in AIF follow-up through the No Findings Left Behind initiative Aims to reduce findings lost to follow-up from the industry norm of 20–40 percent to under 5 percent Strengthens metrics including length of stay, risk adjustment factor (RAF), and hierarchical condition category (HCC) reporting Scalable beyond 1 million advanced imaging reports (CT, MRI, PET) per year "The first-of-its-kind solution we developed with Quantiphi at the University of Miami, Hurricode, will particularly improve quality and safety, identify actionable findings in at-risk patients, and enable other proactive measures that we feel will profoundly contribute to the rapidly expanding field of preventive radiology." — Dr. Alexander M. McKinney, chair of the Department of Radiology at the University of Miami Figure 3: Hurricode AI Platform: Intelligent Radiology Findings and Clinical Integration Hub What's next In the next phase, UHealth plans to automate the transfer of key EHR data into required documentation for more timely and compliant processes. The team also plans to expand Hurricode beyond radiology into pathology, interventional neurosurgery, and cardiology. To explore how generative AI on Amazon Bedrock can transform your organization's workflows, connect with Quantiphi or visit the AWS Generative AI Innovation Center. About the authors Figure 4: Professional Headshot - Giorgia Rematska Giorgia Rematska, PhD Giorgia Rematska, PhD, is a principal architect and machine learning specialist at Quantiphi with over 7 years of expertise in traditional ML, deep learning, and generative AI. She has led initiatives including automated medical document processing, ICD-10 code prediction, and model distillation for privacy-preserving NLP. Figure 5: Professional Headshot - Rakesh Raghu Rakesh Raghu Rakesh Raghu is a senior partner solutions architect at AWS who helps AWS Partners design and build scalable, more secure cloud solutions for public sector customers. He specializes in cloud networking and connectivity and works on migrating workloads to AWS and architecting generative AI solutions. Figure 6: Professional Headshot - Shane Knisley Shane Knisley Shane Knisley is a partner solutions architect with AWS Worldwide Public Sector (WWPS) who helps partners and public sector customers design more secure, compliant workloads across AWS commercial and government Regions. He has over 20 years of IT and cybersecurity experience with deep expertise in RMF and FedRAMP processes.$

How UMiami and Quantiphi optimized radiology coding with Amazon Bedrock

The University of Miami Health System (UHealth) set out to close that gap in both downstream and upstream directions. Working with Amazon Web Services (AWS) and AWS Partner Quantiphi, the UHealth Department of Radiology built a generative AI coding solution on Amazon Bedrock that pairs AI with a radiologist attestation workflow, delivering clinical-grade accuracy while replacing error-prone manual processes.

Controlling delegation in agentic AI with Amazon Bedrock Agent Core

This post introduces a delegation model that gives architects a structured way to assess whether agentic systems meet their authorization and access control requirements. It also shows how Amazon Bedrock Agent Core enforces those boundaries in practice.

How Pariveda built real-time clinical voice notes for Henry Schein One using Amazon Nova and Amazon Bedrock AgentCore

In this post, we show how Pariveda and Amazon Web Services (AWS) worked with HS1 to build real-time voice documentation for dental visits. The solution uses Amazon Transcribe for streaming speech recognition and Amazon Bedrock for structured note generation and orchestration. We also explain the architectural choices required to support real-time performance, template-driven output, security controls, and production scale.

Transforming tactical communications with AWS Wickr and agentic AI

In this post, we explore how Amazon Web Services (AWS) Wickr and AWS WickrGov serve as the nexus for field operators to use well-architected mission workloads and state-of-the-art models on AWS with existing military equipment.

Reducing documentation drift with Amazon Bedrock AgentCore

To address this challenge, Amazon Web Services (AWS) Advanced Partner Corley built an agentic solution for Eutelsat, one of the largest satellite communications operators globally, that automatically synchronizes technical documentation with the source code of in-house projects.

Building supply chain multi-agent workloads in AWS GovCloud (US)

This post shows how to build that on Amazon Web Services (AWS) using Amazon Bedrock, deployed in AWS GovCloud (US). You’ll deploy a working multi-agent workload, see how a supervisor coordinates specialized agents through the Converse API in Amazon Bedrock, and learn which AWS GovCloud (US) details break patterns copied from commercial Regions.

Human-in-the-loop claims processing with Amazon Bedrock AgentCore

In this post, we explore a human-in-the-loop (HITL) framework in the form of architectural patterns teams can use to introduce agentic AI incrementally into eligibility workflows that are adapted to their programs, policies, and risk tolerances.

← Older posts

Newer posts →